Grizzly bear (Ursus arctos) COSEWIC assessment and update update status report: chapter 7

7. Limiting factors and threats

7.1 Abundance

Grizzly bear populations can be affected through direct mortality, or through factors that influence vital rates such as natality. In most Canadian populations, direct human-caused mortality figures largely in the potential persistence of grizzly bears. Natural mortality occurs in all bear populations, and can be substantial, but in nearly all regions including some protected areas, most grizzlies die from human-related causes (Schwartz et al. in press).

Humans kill grizzly bears in a number of ways. All provinces and territories where grizzly bears persist manage the species as a game animal. Hunting seasons are provided for First Nations, resident and, in some cases, non-resident hunters. Bears are commonly attracted to sites of human activity, and may be destroyed as perceived or real threats to life or property. Grizzly bears are shot illegally, perhaps in cases where they were mistaken for black bears, or for malicious reasons. Finally, like all wildlife, grizzly bears are susceptible to accidental human-caused mortality such as collision with vehicles and trains.

In the interior mountains of southern Alberta and British Columbia (BC), and northern Montana, Idaho, and Washington, humans caused 77% of known mortalities of radiocollared grizzly bears, or 85% if suspicious deaths were included (McLellan et al. 1999). Of 83 mortalities, 14 were natural, 16 were legally harvested, 21 were killed in defence of life or property (DLP), 19 were poached, 8 were other human-caused kills, and 5 died of unknown causes.

Benn (1998) reviewed records of grizzly bear mortalities in the Central Rockies Ecosystem (CRE) of Alberta and BC (there is some overlap between the datasets of Benn [1998] and McLellan et al. [1999]). For his whole study area, humans caused 627 (98%) of all documented mortalities. On Alberta provincial lands, 190 human-caused deaths were recorded during 1972-1996. Legal hunters (including First Nations) killed 107 (56%), poachers killed 31 (16%), 48 (25%) were DLP kills, and 4 bears (2%) died of other human causes. Overall mortality rate was estimated at 6.1-8.3%, with a substantial regional disparity. For the Bow River Valley and south, mortality rates were estimated at 1.5-3.1%, whereas north of the Bow River, mortality rates of 7.3-15.9% were estimated.

In the East Kootenays (BC) portion of the CRE, 319 man-caused mortalities were recorded between 1976 and 1996 (Benn 1998). Licenced hunters (excluding First Nations) killed 257 (81%), poachers killed 11 (3%), 48 (15%) were killed in DLP, and 3 (1%) deaths were of unspecified human causes. Man-caused mortality rates were estimated at 1.4% (1976-1981), 2.9% (1982-1996), and 2.5% for the entire period.

In Kootenay, Yoho, and Banff National Parks, humans caused 118 (91%) of 129 known grizzly bear mortalities from 1971 to 1996 (Benn 1998). Of these, 85 (72%) were DLP kills, 22 (19%) were killed on highways or railroads, and 11 bears died of other human causes. The man-caused mortality rates were estimated at 8-10% during 1971-1983, 2% during 1984-1996, and 4.5-5.7% overall. Clearly, even in National Parks which prohibit hunting, grizzly bears are not secure from human-caused mortality.

The data reviewed by Benn (1998) were based on records of reported mortalities, which are strongly biased toward legal anthropogenic sources. Natural mortalities are more difficult to document, especially without radiotelemetry. Therefore, the observations summarized above are best viewed as an evaluation of human-caused mortalities.

A second major problem for grizzly bear managers is the prevalence of undocumented human-caused mortality. In all jurisdictions, grizzly bear kills are subject to compulsory reporting. Most agencies attempt to account for unreported mortalities in their grizzly bear management plans, but documentation, especially for illegal kills, is difficult. McLellan et al. (1999) determined that without radio monitoring, 46-51% of all mortalities of radiocollared grizzly bears would have been unrecorded. For only human-caused deaths (including suspicious, unknown-cause deaths), 34-46% of mortalities would have been undocumented without radiotelemetry. Unreported mortalities have also been documented in an ongoing research program in west-central Alberta (G. Stenhouse, pers. commun.). In addition to illegal kills, it can be difficult to document accidental deaths (e.g., roadkills) and First Nations kills, which are not subject to compulsory reporting in all jurisdictions.

Management responses to grizzly bear/human conflicts often include capture and translocation rather than destruction of the offending bear (Schwartz et al. In press). Although translocated bears have at least a chance at survival not realized if they were destroyed, translocation should not be considered a solution to conflicts. Homing ability is well-developed in grizzly bears, and many bears return quickly (Miller and Ballard 1982). In the Yellowstone area, survival rates of transported bears were lower than for bears that were not moved (0.83 versus 0.89), especially for males and adult females (Blanchard and Knight 1995). In northwestern Montana, 38% of translocated bears died within 2 years (Riley et al. 1994). Irrespective of survival of the transported individual, any successful translocation must be considered functionally equivalent to a mortality to the source population.

Because grizzly bears are long-lived, have low reproductive potential, and their populations are difficult to monitor, they are extremely difficult to manage. Current demography in a grizzly bear population is the consequence of a series of events that occurred over the previous 10-20 years (Doak 1995). Managers need to ensure that population age distributions are “healthy”, and that reproduction and recruitment are maintained. Excessive mortality and disruption of breeding potential can lead to mistaken impressions of viability, wherein a few old bears persist at very low densities over several years. Such “living dead” populations exist in France (Camarra 1999), northern Italy (Osti 1999), and Spain (Clevenger et al. 1999).

Population Viability Analysis (PVA) is a technique used to predict population persistence. Increasingly, it is used to model extinction risk for grizzly bear populations (Mills et al. 1996; Herrero et al. 2000). Input variables for PVA are region-specific, so reasonable estimates of bear population characteristics and habitat conditions must be known and foreseeable. Because of the range of those variables among grizzly populations, it is not usually feasible to apply these models to predict the extinction risk in arbitrary areas, or to determine minimum sizes of protected areas (Mattson et al. 1996). Use of PVA to influence management actions requires at least 2 difficult decisions: what probability of extinction is acceptable, and for what length of time does an acceptable risk of extinction need to apply? Reducing the probability of extinction of a grizzly bear population to zero is impossible given existing conditions and stochastic events. Managers who need to maximize the likelihood of population persistence must be prepared to aggressively minimize, and perhaps reverse, human influences on grizzly bear habitat and populations. As willingness to accept risk of extinction increases, regulated limits on human activities can relax. Similarly, ensuring persistence over a 1000-year period requires far more restrictive planning than if the acceptable time frame were 100 years (Mattson et al. 1996). Such decisions have profound bearing on human society, and cannot be made by wildlife managers alone.


7.1.1 The effects of hunting on grizzly bear populations

All provinces and territories with grizzly bears provide hunting seasons for them. Licenced hunting accounts for about 84% of documented man-caused mortality (Table 11). Using mean annual harvests during 1990-1999 (Table 11) and the current population estimates (Table 6), annual harvests have been 0.6 – 3.4%. Undoubtedly, harvests have exceeded these averages in some years and in some areas.

Grizzly bear populations “under optimal conditions for reproduction, natural mortality, and with males twice as vulnerable as females” are estimated to be able to sustain a maximum annual harvest rate of 5.7% (Miller 1990b:357). Grizzly bear management strategies in Canadian jurisdictions include goals of total man-caused mortality of 6% or less of estimated populations. Generally, other man-caused mortalities are subtracted from total quotas before harvest allocations are made. Assuming that population estimates are accurate and all man-caused mortalities are documented, hunting mortality in Canadian jurisdictions is likely to be sustainable. Unfortunately, the fact that neither assumption is reasonable dictates that a conservative approach to all parameters is required.

As with all polygynous species, more male grizzly bears can be harvested than females without detriment to the population (Caughley and Sinclair 1994). Most management agencies actively direct harvest toward male bears, by protecting family groups and by scheduling hunting seasons when males are relatively active. In addition, male bears are larger and are preferred by most hunters. During 1990-1999, 65% of annual harvest in Canada has been male (Table 11). Consequently, the reproductive core of a bear population, the adult females, is granted greater protection. However, this harvest strategy, or any strategy that concentrates relatively high harvest on males, may exert pressure on the population against the natural density of males and adult sex ratio toward which the population gravitates under unharvested conditions. The social and behavioural implications of this are wholly unknown, but are likely important and warrant further study.

Heavy hunting pressure can alter grizzly bear population characteristics. An estimated annual average harvest rate of 11% over 12 years reduced an Alaskan bear population by 36%, and the female population by 32% (Reynolds 1999). Following 10 years of deliberate effort to reduce the bear population in another Alaska study, the sex ratio declined from 82 males/100 females to 28 males/100 females, although no change in density was measured (Miller 1995).

Information on sex and age composition of the harvest should not be considered a proxy for those parameters in the population, because different classes of bears may be more vulnerable to hunters (Bunnell and Tait 1980) and because hunters are selective (Miller 1990b). Interpretations of harvest data also often assume a stable age structure in the population (Miller 1990b). Misinterpretation of data on harvest characteristics may lead to unsubstantiated conclusions of population status. For example, commonly used indicators of overharvest include a declining harvest age structure or an increase in proportion of females in the harvest (Miller 1990b), so failure to detect these trends might be interpreted as an indicator of population stability. However, there was no change in population age distribution in a heavily hunted area in Alaska (Miller 1995). Similarly, the relative vulnerability of male bears to harvest (Bunnell and Tait 1980) may conceal their actual decline in an overharvested population. In addition, low harvest rates mean small sample sizes, with correspondingly low statistical power to detect population decline from harvest data (Harris and Metzgar 1987).

Protection of female bears is not assured by establishing minimum proportions of males in the harvest (Miller 1990b). If those proportions are not met, restructuring of regulations to increase the male harvest (e.g., earlier season opening) may be less appropriate than decreasing the female harvest. It is more important to protect an absolute number of female bears than to maintain a particular proportion of females in the harvest.

Trophy hunting of male bears is considered by some to be neutral or beneficial for populations because reduced male density is believed to increase cub production and survival, and to thereby stimulate population growth (review in Miller 1990a). The purported mechanism is that adult males are infanticidal and suppress population growth, and their removal is compensated by reduced intraspecific stress and increased recruitment. However, a review of research studies was inconclusive as to evidence of such density-dependent compensation in grizzly bear populations (McLellan 1994), and it was recommended that until such evidence was clear, managers should presume that rate of recruitment will not increase as a result of reduced population size (Taylor 1994).

Recent research suggests the opposite effect. In Sweden, brown bear cub survival was lower in an area with higher adult male mortality, and immigrating males were implicated as the cause of cub deaths (Swenson et al. 1997; In Press). Cub survival was reduced for 1.5 years after adult males were removed, indicating social disruption persisted for that long. When no adult males were removed for at least 1.5 years, cub survival was 0.98 to 1.00, suggesting that established resident males killed few cubs. Swenson et al. (1997) concluded that killing 1 adult male bear had a population effect equivalent to removing 0.5 to 1 adult female. In comparing hunted and unhunted grizzly populations in Canada, and controlling for differences in habitat quality and density, Wielgus and Bunnell (1995; 2000) found lower reproduction rates, mean litter size, and age at first parturition in the hunted population. Males immigrating to replace hunter-killed males were considered potentially infanticidal, and resident females avoided those bears and the high-quality habitats they used. The polarization of opinion on this topic among researchers, compounded by a dearth of conclusive data, supports the recommendations that this issue be treated conservatively and become the focus of directed research.

Because of the difficulty in obtaining accurate demographic data, a lack of understanding of the full suite of consequences of bear mortality, and the species’ inherent low ecological resiliency (Weaver et al. 1996), management plans and harvest goals must in all cases be cautious and conservative.


7.1.2 The trade in bear parts

Asian medicine has relied on bear parts for thousands of years. Today, demand for bear parts persists with practitioners of traditional medicine, has expanded from China to Korea and Japan, and has followed Asian immigrants to other continents. Bile from bear gall bladders has been the substance most widely sought, but markets exist for other body parts, especially paws. Bear bile and galls comprise putative remedies for a number of internal ailments including diseases of the liver, heart, and stomach.

Bear bile and galls are valuable. Documented retail prices can reach US $500/gram for bile, and US $2,000 for whole gall bladders (Servheen 1999b). Reliance on bears and bear parts has contributed substantially to the decline in distribution and populations of several bear species in Asia. Exploitation of bears for medicinal supplies has expanded to include North American bear populations to address demand of traditional users in both North America and Asia. All bear species are used to provide medicinal ingredients. Because of the difficulty in successfully prosecuting offenders, and the relatively mild penalties imposed by many jurisdictions, poaching for wild bears is perceived to be highly profitable.

With the exception of a limited market for pelts, no Canadian jurisdiction permits trade in grizzly bear parts. Therefore, all such trade is illegal and impossible to document or monitor, and it is difficult to evaluate the magnitude of exploitation of North American bear populations. However, reports of successful prosecutions (e.g., BCMOE 2001a) indicate that it occurs. Efforts to curtail trafficking in bear parts have improved in some areas. For example, BC passed legislation in 1997 that prohibits possession of bear gall bladders or any part or derivative of a bear gall bladder. Further, it is illegal in BC to possess any product that contains--or is alleged to contain--bear bile. Although aggressive legislation is essential to inhibit trade in bear parts and undoubtedly is a partial deterrent, offences and prosecutions continue.

As populations of wild Asian bears continue to drop from habitat degradation and excess killing, the Asian supply of bear parts will decline further and pressure to compensate for this shortfall by exploiting other bear populations will increase. There are more wild bears in North America than in the rest of the world combined (Servheen 1999b). In light of the high profits available from trafficking in bear galls, bile, and other parts, it is probable that North American bear populations, including grizzlies, will come under increasing pressure to supply this market.

Medicinal use does not account for all trade in bear parts. The trophy value of grizzly bears, in particular, also inspires some degree of poaching and commercial traffic in grizzly trophies (e.g., BCMOE 2001b). If legal hunting opportunities for grizzly bears in Canada and elsewhere become more restricted, this threat can be expected to increase.

7.2 Habitat

Habitat perturbations influence an area’s capacity to support grizzly bears. Although natural and anthropogenic habitat alterations can be beneficial to bear populations (e.g., enhancement of early forest successional stages through fire or timber harvest), of greater concern to grizzly bear status and conservation are those activities which degrade habitat effectiveness. Foremost in importance among habitat alterations are those which convert grizzly bear habitat to areas which will not be suitable for bears either permanently or over a long-enough term to affect population characteristics. Included in this category are certain resource-extraction industries, agriculture, and residential development. For many years, such developments proceeded throughout much of grizzly bear range effectively or completely unmitigated. Recently, however, in response to acknowledged declines in the global, North American, Canadian, and local distributions of grizzly bears, proposed developments are increasingly subject to critical scrutiny. Examples of recent attempts to assess the effects of proposed industrial developments on grizzly bear habitat and populations include Herrero and Herrero (1996), Diavik (1998), and BHP (2000).

Mining and hydrocarbon extraction are of concern because the nature of valuable geologic deposits dictates that mines are dug and wells are drilled where the deposits happen to be found; locating and extracting those resources in other, more environmentally appropriate locations is often not possible. Society’s demand for those resources, therefore, directs that within certain constraints, we will have those mines and wells and some degree of habitat loss will occur. Particularly precious resources such as oil and gas can drive economies on the provincial or federal scale, exerting considerable pressure against the need to preserve grizzly bear habitat.

For example, gross revenues from hydrocarbon production in Alberta exceeded $26 billion in each of 1996 and 1997, and 76% of production is exported from the province (Alberta EUB 1999). In support of this, over 200 000 wells have been drilled in Alberta since 1902, with annual increments of 8 000 to 13 000 during recent years. In addition, by the end of 1998, the total length of pipelines within Alberta was 264 000 km; the oilfield road network is thousands of kilometres long. A substantial, but unknown, fraction of these wells, pipelines, and roads occurs within current grizzly bear distribution. Each well site, access road, and servicing pipeline constitutes a long term or permanent habitat alteration, and most should be considered negative. Pipelines and roadside verges may provide foraging opportunities for bears (Nagy and Russell 1978), but increased vulnerability to hunters may offset any potential advantage. Hydrocarbon exploration and development are also progressing rapidly in southwestern Northwest Territories (NWT), and interest in development of a pipeline along the length of the Mackenzie Valley has been recently renewed.

Until the late 1980s, the grizzly bears of the Canadian Arctic were relatively remote from industrial developments. However, the announcement of the discovery of diamonds in 1991 triggered unprecedented interest and exploration activity. Between 1991 and 1993, more than 23 000 diamond-related mineral claims, encompassing over 160 000 km2, were staked in the Slave Geologic Province of NWT and Nunavut (Mining Recorder’s Office, Department of Indian Affairs and Northern Development, Yellowknife, unpublished records). By the end of 2001, one diamond mine was in production, a second was nearing start-up, and several more were in advanced stages of exploration, or development applications were in review. In 1999, the sole producing diamond mine contributed 19% of the Gross Domestic Product of the Northwest Territories (BHP 2000). Development of mines on the Arctic tundra brings a new threat to grizzly habitat. Although mine footprints are relatively small, habitat effects are exacerbated by the open tundra landscape, the local intensity of disturbance, the relative scarcity and importance of high quality habitat patches, and exceptionally low bear density. Each development also serves as a potential site of conflict between bears and humans, with associated risk of bear mortality. Because of the enormous home range sizes typical of Arctic grizzlies (Table 5), bears have an elevated risk of encountering even widely dispersed and low-density developments, especially if those sites are attractive to bears because of the presence of anthropogenic food material or bears’ natural curiosity and inclination to investigate potential food sources. The birth and rapid development of the diamond-mining industry in Canada point to the capriciousness and unpredictability of resource extraction activities and the markets that drive them. Barely one decade ago, few would have anticipated intensive interest in mining within the barrenlands of the central Canadian Arctic. It is conceivable, and even predictable, that within the next decades, pressure will mount to exploit resources which have not yet been discovered, or for which current demand is non-existent.

Commercial timber harvest in Canada alters a substantial amount of grizzly bear habitat each year. During 1997, clearcut logging totalled 175 808 hectares in British Columbia, 50 697 hectares in Alberta, and 429 hectares in NWT (BC MOF 1998a; NRC 1998). In the Yukon, 1921 hectares were clearcut in 1996. Habitat effects of timber harvest are dynamic, and depending on post-harvest treatments, bears may respond positively to early seral stages during revegetation of cutblocks. However, McLellan and Hovey (2001a) found very little bear use of large regenerating cutblocks in southeastern British Columbia, because few bear foods occurred there. With all cutblocks, for at least a short term after logging, habitat effectiveness is profoundly reduced. Associated with timber harvesting is the development of roads. As of 1998, the BC provincial forest was accessed by a network of roads including 43 000 km of Forest Service Roads, 120 000 km of permitted (operational) roads, and 150 000 km of abandoned roads and trails, for a total of 313 000 km (BC MOF 1998b). “Abandoned” roads and trails are no longer maintained, but in most cases are likely accessible to all-terrain vehicles. Each year, 5 000-10 000 km of roads are added to this total.

Agricultural development was probably responsible for a substantial component of grizzly bear range contraction in Canada, and continues to date. Conversion to crop land permanently deletes that land as grizzly bear habitat. Livestock grazing leads inevitably to grizzly bear mortality when bears are removed because of real or perceived threats of depredation (LeFranc et al. 1987).

Of all anthropogenic habitat alterations within grizzly bear range, the most disruptive is probably residential development. With increasing affluence, more people build homes on the fringes of grizzly bear distribution. With most industrial developments, human activities are confined temporally on a diurnal, seasonal, or rotational scale. When the people leave, so does the habitat disruption, with the possible exception of a disturbed footprint. Residential developments are more disruptive because the human presence is virtually continuous and permanent. Although the area of habitat displacement related to a single home may be small, each contributes to the cumulative influence of whole subdivisions, and works in concert with other developments and activities in the region. Additionally, the attractants usually associated with human homes (e.g., garbage, pet food, livestock) dictate that bears with home ranges overlapping with permanent human habitation are at extremely elevated risk of mortality (McLellan 1994).

In most cases, effects of individual human activities do not operate in isolation to influence grizzly bear habitat or populations. For example, in west-central Alberta, human activities including timber harvest and coal mining apparently reduced grizzly bears over the period 1971-1995 (Herrero and Herrero 1996). This apparent population decline could not easily be attributed to specific causes, but a combination of excessive human-induced mortality, and habitat loss and alienation due to development were probable causal factors. Overall, existing human developments and activities in the area appeared to have been the primary factors leading to apparent carnivore population declines. Technological advances and enhanced modelling power (e.g., Geographical Information Systems) have improved the ability of managers to predict, evaluate, and mitigate cumulative environmental effects, but the rate at which cumulative effect scenarios are developing in grizzly country is increasing rapidly.

Because it is not feasible to quantify all contributions to grizzly bear habitat alteration, and recognizing that they are a direct consequence of human population, the trend in numbers of people is a useful analogue of these activities. Strong associations between human density and loss of carnivore populations have been documented (Woodroffe 2000). In addition, the distance to, and size of, human population centres were strongly correlated with grizzly bear habitat effectiveness (Merrill et al. 1999). Human density in Canada increased almost 9-fold between 1861 and 2000, the period of most decline in the abundance and distribution of grizzly bears. Although the bulk of Canadians continue to live in the east, the rate of population growth in recent years is higher in the provinces and territories that have grizzly bears. Between 1971 and 2000, the Canadian population increased by 43%, less than the increase in Alberta (84%), BC (86%), Yukon (61%), or Northwest Territories including Nunavut (101%). The total population for the region increased by 85% over the same period. Population growth rates may decline over the short-term future, but absolute population within grizzly bear distribution will certainly increase (McLellan 1994; Herrero et al. 2000).

There is a clear link between habitat degradation and population effects in grizzly bears. Doak (1995) modelled the reduction of habitat quality in the Yellowstone area and predicted that even small amounts of habitat degradation could result in rapid declines in grizzly population growth rates. Even more insidious was his finding that when the rate of degradation was slow (1% per year), it could take more than 10 years to detect critical amounts of degradation beyond which bear populations could begin long-term declines.


7.2.1 Potential consequences of climate change on grizzly bear habitat

Global warming could lengthen the growing season particularly for bears at high latitudes, increasing the period during which green forage is available. These effects could be direct, in providing more vegetation for bears to consume, and indirect, in increasing habitat quality for bear prey--if those enhanced resources would be present at times when they are accessible to bears or migratory bear prey. This could shorten the duration that bears are confined to their dens and in a negative energetic state. If bears are able to exploit enhanced food resources, conceivably this could result in larger bears, lower mortality rates, and higher litter sizes and other reproductive parameters. Some landscapes may increase in overall productivity, and bear carrying capacity could increase. These changes may be most noticeable in marginal current bear habitats such as Arctic or alpine tundra, and if productivity of high-Arctic environments increases, grizzly bear distribution may expand to include those areas.

Conversely, increasing temperatures will raise sea level and result in the inundation and loss of some of today’s most productive coastal bear habitats. If warming is accompanied by generally drier conditions, plant-community structure may be altered such that productive, moist environments decline and are replaced with poorer-quality assemblages, with less palatable or less nutritious vegetation, and lower biomass of potential bear prey. Increasing temperatures may also facilitate human habitation of areas presently considered inhospitable, with the attendant problems related to conservation of grizzly bears within human-occupied landscapes.

However, such predictions are simplistic and wildly speculative because of the complexities of interactions among components of grizzly bear habitat. It is impossible to model the effects on each trophic element below grizzly bears, and many of those elements have keystone roles in defining bear habitat quality. For example, accurately predicting the positive or negative consequences of climate change on salmon or caribou life history and populations is unrealistic.

Throughout Section 7 of this report, evidence is presented that the greatest threats to the conservation of grizzly bear populations and habitats are those posed by human activities. Among all prophecies related to climate change, none is more unpredictable than human responses to a changing climate. How these may influence grizzly bears is beyond speculation.

7.3 The effects of roads on grizzly bears

The U.S. Fish and Wildlife Service (1993) believes that “roads probably pose the most imminent threat to grizzly habitat today” (pg. 21), and that “the management of roads is the most powerful tool available to balance the needs of bears and all other wildlife with the activities of humans” (pg. 145). Although direct mortality of grizzly bears from roads (i.e., roadkills) has been documented, the most important effects of roads on grizzly bears are (1) loss of habitat effectiveness because of bears avoiding the disturbance associated with roads, and (2) shooting mortality facilitated by the development of new access routes for hunters and others with firearms.


7.3.1 Habitat effects

Grizzly bears may be vulnerable to individual disruption arising from construction, maintenance, and use of linear developments. Efficient foraging strategies of bears were disrupted near human facilities including roads in Yellowstone National Park (Mattson et al. 1987). Archibald et al. (1987) documented, between prehauling and posthauling, a 33% and 39% reduction, respectively, in the number of times that 2 bears crossed a logging road in the Kimsquit Valley in British Columbia. These bears did not appear to habituate to logging traffic after 2 years of hauling. Grizzlies in southern Alberta did not appear to habituate to high-speed, high-volume traffic on the Trans-Canada Highway (Gibeau et al. 2002). However, some authors believe that grizzly bears may become accustomed, or desensitized, to predictable occurrences, including traffic (Tracy 1977; Bader 1989; McLellan and Shackleton 1989). Presumably, bears are unlikely to habituate to infrequent traffic, and individuals may react more vigorously to once-per-week vehicle passages than to vehicles passing every few minutes. Similarly, regular spacing of vehicles is likely to contribute more toward habituation than the same volume of traffic concentrated in a brief period. Habituation may permit some bears to exploit high-quality habitats adjacent to roads; however, it may also greatly increase the likelihood of collision mortality or negative bear-human interactions, with the attendant risk of management action to remove problem bears. Another factor likely to influence bears’ responses to human activity is whether or not bears are hunted. Habituated bears do not survive in hunted populations.

Disturbance along roads may result in habitat avoidance for grizzly bears. Logging-truck traffic in the Kimsquit Valley in British Columbia resulted in a 78% reduction in use of the “Zone of Hauling Activity” by radiocollared bears compared to non-hauling periods (Archibald et al. 1987). For 14 hours/day, 3%-23% of each bear's home range was unavailable to them because of disturbance. Because bears used these areas when hauling was not going on, it was clear that these areas were of value to the bears. In rich habitats such as coastal BC, where bear home ranges are small, these losses can limit access to important food sources.

In southeastern British Columbia, McLellan and Shackleton (1988) calculated that 8.7% of their total study area was effectively lost to bears as a result of road avoidance. Mattson et al. (1987) estimated that habitat effectiveness lost to developments was sufficient to support 4-5 adult female grizzly bears in their study in Yellowstone National Park.

On the Rocky Mountain Front in Montana, Aune et al. (1986) reported that for all monitored bears, “In spring and fall the 0-500 m distance to road category was used significantly less than expected. All other categories were used as much as expected when compared to random chance. In summer this distance category was used as much as expected. Results imply that in summer for all grizzlies sampled, road influence zone could be less than 500 m but during spring and fall may be at least 500 m.” (pg. 59). Road-habituated bears “showed no significant road avoidance in spring or summer in the 0-500 m category. However (they) did significantly avoid this zone in fall. It appears that any road influence on these bears would be less than 500 meters from the roadside for spring and summer.” (pg. 62). Bears which were classed as non-habituated to roads within their home ranges “showed significant avoidance of the 0-500 meter road category for all three season(s) and for fall the avoidance was significant to 1000 meters of the roads.” (pg. 62). In northwestern Montana, grizzlies used habitats within 914 m of roads at just 20% of the predicted rate, and used areas >1860 m from roads more than predicted (Kasworm and Manley 1990). Female grizzlies avoided the Trans-Canada Highway in Banff National Park irrespective of habitat quality; male bears also avoided the highway, except when it traversed high-quality habitats (Gibeau et al. 2002). In southcentral BC during spring, 85% of bears avoided habitats--including highly-preferred habitats--adjacent to transportation corridors including the Trans-Canada Highway and a trans-continental railroad (Munro 1999). Avoidance was most pronounced in female bears, which avoided transportation corridors in all seasons.

Ruediger (1996) hypothesized that net impact on carnivores increases with construction standard of roads. High-speed, high volume interstate highways probably have a greater impact on carnivore populations than do small rural roads. Intuitively, heavily used roads probably have a larger negative effect on grizzly bears than quieter roads. Gibeau (2000) reported that adult female grizzly bears that would not cross the Trans-Canada Highway in Banff National Park (21 000 vehicles per day with an average speed of 110-115 km/hr), would cross other 2-lane highways (2230 – 3530 vehicles per day with an average speed of 80-115 km/hr). In northwestern Montana, the number of grizzlies showing selection for 500-m buffers surrounding roads decreased as traffic volume increased (Mace et al. 1996). All bears in this study avoided buffers around roads with >60 vehicle passes per day, and most avoided buffers around roads with >10 vehicle passes per day, but there was some selection, or neutrality, for buffers surrounding roads with <10 vehicle passes per day.

“Trails” include foot, bicycle, and equestrian trails, and may include roads that are closed to public use. Trails used by motorized off-highway vehicles are presumed to have the same effects on grizzly bears as roads. In northwest Montana, grizzlies avoided habitats within 274 m of trails (Kasworm and Manley 1990). Overall, trails displaced grizzly bears less than roads did in that study. In the Swan Mountains, Montana, grizzlies were found significantly further than expected from trails during spring, summer, and autumn (Mace and Waller 1996). These authors concluded that grizzly bears using the hiking area have become negatively conditioned to human activity occurring within and outside the area, and that they minimized their interaction with recreationalists by spatially avoiding high-use areas.

Roads and other linear developments may serve either as filters or barriers to the movements of grizzly bears. A highway appeared to exert short-term deflections on movements by 3 bears, all adult females, in Alaska (Miller and Ballard 1982). Gibeau et al. (2002) reported that in 7 years, a single radio-collared adult female grizzly and 2 radio-collared adult males crossed the Trans-Canada Highway in Banff National Park, and concluded that the highway was a barrier to adult female bears. Based on genetic sampling, Highway 3 through the Crowsnest Pass in southern Alberta and British Columbia is nearly a barrier to female grizzlies, and has apparently reduced male movement as well (Proctor et al. In Press). In Slovenia, a highway served as a home range boundary for 3 radio-monitored adult brown (grizzly) bears (Kaczensky et al. 1994). These bears approached the highway, closely at times, but the 2 females did not cross it and the male crossed it only twice.

No absolute threshold has been determined to define a road density which is acceptable to grizzly bears. In the Swan Mountains of northwestern Montana, grizzly bears used only areas with total road density (including closed and rarely used roads) <6.0 km/km2 (Mace et al. 1996). Merrill et al. (1999) provide evidence for 1 km/km2 of roads and trails as being a broader-scale threshold for relatively productive habitats such as those found in the US Selkirks and Cabinet-Yaak ecosystems. At some density, roads will become complete barriers or mortality sinks to grizzlies (Ruediger 1996), even if adjacent habitats would support their populations. However, it is difficult to predict the consequences of any particular road density on a bear population since many factors such as habitat, road type, and traffic volumes also affect the degree to which bears avoid roads.

Effort has been made to standardize allowable road densities in grizzly bear recovery zones. At present, these range from 0.75 mi. open road/mi2 (0.47 km/km2) to 1.0 mi. open road per mi2 (0.62 km/km2) (US Fish and Wildlife Service 1993). The Gallatin National Forest in Montana has adopted an open road density standard of 0.5 mi./mi2 (0.31 km/km2) (Paquet and Hackman 1995). It should be noted that the definition of “open roads” includes roads which are closed to public users but which are subject to administrative use exceeding “…one or two periods that together … exceed 14 days during the time bears are out of the den (usually between April 1 and November 15)” (US Fish and Wildlife Service 1993:148).

Not all authors agree with these standards. Craighead et al. (1995) argue that these densities are much too liberal. They advocate, for grizzly bear recovery and conservation, an open road density no higher than 1.0 km per 6.4 km2 (0.16 km/km2; 0.25 mi/mi2). They further recommend that roads on federal or state land that exceed this density be closed and obliterated.

Social disruption of grizzly bear populations resulting from linear developments has also been reported. Most records of habitat avoidance (see above) probably also represent cases of social disruption because displaced bears are forced into concentrations higher than those they might naturally seek. If different cohorts of bears demonstrate different tolerance for disturbance, then the resulting spatial arrangement of bears may also be suboptimal. Subordinate cohorts of bears were displaced into poorer-quality habitats near developments by more dominant classes, particularly adult males, in Yellowstone National Park (Mattson et al. 1987). McLellan and Shackleton (1988) also determined that adult males used remote areas whereas adult females and some subadults used areas closer to relatively low-use roads in southeastern BC. Conversely, female grizzlies remained further than males from high-volume highways in southwestern Alberta, regardless of relative habitat quality (Gibeau 2000).


7.3.2 Population effects

Many authors have reported mortality in grizzly bear populations as a direct or indirect consequence of linear developments. Grizzly bears may be killed in collisions with vehicles (LeFranc et al. 1987; Gibeau and Heuer 1996). Gunson (1995) analyzed records of 798 grizzly bear mortalities on provincial lands in Alberta from 1972 to 1994; 5 bears were killed by trains, and 4 by other vehicles. Although such mortalities can be important to small or low-density populations, most authors concur that greater mortality effects arise out of indirect consequences of the construction of roads and other linear developments.

During winter (roughly November 15 to April 1), nearly all grizzly bears are in dens (Linnell et al. 2000). Bears may be displaced from their dens by industrial activity or other disturbance (Harding and Nagy 1980; Swenson et al. 1997). Bears that flee their dens during winter will likely experience severe physiological stress and may die. Pregnant females may lose their cubs (Swenson et al. 1997), and abandoned cubs will not survive. The danger of winter industrial operations within grizzly bear denning areas is that precise locations of dens will not be known, and new construction or other activities may inadvertently approach them very closely.

Linear developments like roads generally lead to increased mortality for grizzlies. Benn (1998) investigated the location of recorded grizzly bear kills in the Central Rockies Ecosystem. In Banff and Yoho National Parks, all 95 human-caused bear deaths with known locations occurred within 500 m of roads or frontcountry (i.e., road-accessible) facilities, or within 200 m of trails or backcountry facilities. In the Alberta portion of the Central Rockies Ecosystem (CRE), 153 (89%) of 172 known-location kills were within 500 m of a road or within 200 m of a trail. In the East Kootenay (BC) portion of the CRE, 122 (71%) of 172 kills with known locations occurred within 1000 m of a road or trail. Spatial analyses indicated that grizzly kill sites were not random with respect to the occurrence of roads and trails. Again, it must be noted that Benn’s (1998) data pertain to reported mortalities of non-radiocollared bears, and are therefore biased against natural mortalities. The distribution of natural grizzly bear mortalities may be independent of human developments including roads and trails.

In the Flathead Valley of southeastern BC, 7 of 13 successful grizzly bear hunters had been on a road when they shot their bear (McLellan 1989b). Many other authors have identified shooting mortality in grizzly bear populations that was related to roads or other industrial access (e.g., Aune and Kasworm 1989; Horejsi 1989; Knick and Kasworm 1989; Nagy et al. 1989; Titus and Beier 1992). Conversely, in a protected grizzly bear population in northwestern Montana annual mortality rates were 15 times higher in wilderness areas than in multiple-use areas, primarily from self-defence and mistaken-identity shootings (Mace and Waller 1998).

Even in unhunted populations, the geographic location of most human-caused grizzly bear mortalities is strongly correlated to human developments. All 8 human-caused mortalities in a study in northwestern Montana resulted from road access and illegal killing or management-related removals (Mace et al. 1996). Mattson et al. (1996) reviewed grizzly bear mortality in the Greater Yellowstone Ecosystem (GYE). A disproportionate 68% of all mortality occurred in habitat substantially impacted by humans yet this habitat represented 33% of the total habitat available to grizzly bears. Mortality in these impacted habitats was 5.8 and 11 times greater than the lowest rates in United States Forest Service roadless areas and United States Parks Service backcountry, respectively. Doak (1995) estimated that mortality risk for grizzly bears in the GYE was 5 times higher near roads.

Indirect mortality as a result of linear developments may occur in other forms. Mattson et al. (1987:271) stated “¼that avoidance of roads and developments by grizzly bears in Yellowstone Park probably resulted in poorer condition adult females and, consequently, higher mortality rates and lower fecundity for the cohort.” Gibeau (2000) reported that bears living in areas of unrestricted human access used lower quality habitat and travelled more than bears in restricted areas, thereby retaining less energy for growth and reproduction.

Indirect population-level effects may occur at a broader scale as well. That adult female grizzlies never crossed the Trans-Canada Highway in Banff National Park (Gibeau 2000) indicates potential interruption of population connectivity, and raises concerns about resultant effects on genetic diversity within this population. Adult bears rarely crossed a highway in Slovenia, and inbreeding is a concern in this small population of bears (Kaczensky et al. 1994). Reduced litter sizes and other indicators of inbreeding depression have been reported for inbred, captive brown bears (Laikre et al. 1996). Genetic diversity is important in maintaining evolutionary potential and individual fitness. Maintenance of genetic diversity in grizzly bear populations, however, is dependent upon connectivity to populations on the scale of the entire North American distribution (Paetkau et al. 1998). For isolated bear populations such as in Yellowstone and the North Cascades, a near-complete loss of genetic diversity is likely unless connectivity is restored or the population is augmented (Paetkau et al. 1998).

Table 12. Mean annual recorded man-caused mortalities inCanadaduring 1990-1999.
Jurisdiction Hunter killsFootnote a Non-hunting man-caused Total
Males Females U/k sex Total Illegal DLPFootnote b Other
Alberta 9.1 4.7 0.0 13.8 3.7 4.2 3.1 24.8
British Columbia 187.3 101.6 0.0 288.9 5.9 40.2 n/a 335.0
YukonFootnote c 51.0 27.5 0.0 78.5 n/a 13.2 1.5 93.2
NWT and NunavutFootnote c 8.1 1.7 1.0 10.8 n/a 9.4 n/a 20.2
ISR and GSAFootnote d 21.1 5.2 4.1 30.4 included in hunter kills included in hunter kills included in hunter kills 30.4
Total 276.6 140.7 5.1 422.4 9.6 67.0 4.6 503.6

Page details

Date modified: